[
https://issues.apache.org/jira/browse/AVRO-1385?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13792873#comment-13792873
]
Doug Cutting commented on AVRO-1385:
------------------------------------
> Following this principle, shouldn't int be accepted as a reader schema for a
> long? This now seems inconsistent to me.
This would be reasonable if we wanted to check that every value fits in 32 bits
and throw exceptions when they don't. We shouldn't silently discard
information.
> Enum reader/writer resolution doesn't report errors
> ---------------------------------------------------
>
> Key: AVRO-1385
> URL: https://issues.apache.org/jira/browse/AVRO-1385
> Project: Avro
> Issue Type: Bug
> Components: java
> Affects Versions: 1.7.5
> Reporter: Christophe Taton
> Attachments:
> AVRO-1385.base=a56e7f13c4fa976b7926649b058665f42850a349.20131010-182234.patch
>
>
> According to the specification, when a reader enum does not contain all
> symbols from the writer enum, an error should be signaled.
> ResolvingGrammarGenerator fails to report such errors.
> In particular, mkEnumAdjust() silently maps unknown symbols of the writers
> with symbols named "No match for [writer symbol]".
> https://github.com/apache/avro/blob/trunk/lang/java/avro/src/main/java/org/apache/avro/io/parsing/ResolvingGrammarGenerator.java#L409
> Shouldn't this instead create an ErrorAction?
--
This message was sent by Atlassian JIRA
(v6.1#6144)
